A first plurality of data points related to visitors to at least one website is received. The data points comprise at least an identification of the visitor and an interaction of the visitor with the website. A target audience comprising at least some of the visitors having a known, desired interaction and a plurality of selection rules defining tolerances for a similarity audience are received. The selection rules include a correlation score and an audience composition index. A similarity audience from a plurality of clusters defined by a number of unique visitors is selected wherein at least some of the unique visitors share at least one interaction in common, and the similarity audience comprises at least one cluster of the plurality of clusters satisfying the plurality of selection rules. Digital content is generated for electronic transmission to a plurality of computing devices associated with members of the similarity audience.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method, comprising: receiving, for at least one website, a first plurality of data points related to each visitor of a first plurality of visitors to the website, the first plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; receiving, at a computing device, a target audience comprising at least some of the first plurality of visitors having a known, desired interaction; receiving, responsive to a user input to the computing device, a plurality of selection rules defining tolerances for a similarity audience, the plurality of selection rules including a correlation score and an audience composition index; selecting, using the computing device, the similarity audience from a plurality of clusters defined by a number of unique visitors, at least some of the unique visitors sharing at least one interaction in common, and the similarity audience comprising at least one cluster of the plurality of clusters satisfying the plurality of selection rules; and generating, from the computing device and for electronic transmission to a plurality of computing devices associated with members of the similarity audience, digital content.
2. The method of claim 1 , further comprising: generating a cosine similarity index for each cluster of the plurality of clusters using the target audience as input, wherein the correlation score is a minimum value for the cosine similarity index of clusters of the plurality of clusters forming the similarity audience.
3. The method of claim 1 , wherein the audience composition index is calculated according to: I = ( A B ) ( C D ) * 100 wherein I is the audience composition index; A is a number of visitors having the known, desired interaction within the similarity audience; B is a total number of visitors within the similarity audience; C is a total number of visitors having the known, desired interaction within the plurality of clusters; and D is a total number of visitors within the plurality of clusters.
4. The method of claim 1 , further comprising, before receiving the first plurality of data points: receiving, for at least one website, a second plurality of data points related to each visitor of a second plurality of visitors to the website, the second plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; and performing a clustering algorithm that generates the plurality of clusters using the second plurality of data points as input, the second plurality of data points being a sparse data set.
5. The method of claim 4 , further comprising: generating, for a cluster of the plurality of clusters, at least one assignment rule that associates a behavior common to at least some of the unique visitors in the cluster, with the cluster; and assigning each visitor of the first plurality of visitors to at least one cluster of the plurality of clusters, before selecting the similarity audience, using the at least one assignment rule.
6. The method of claim 5 , further comprising: re-assigning at least some visitors of the second plurality of visitors to at least one cluster of the plurality of clusters, before selecting the similarity audience, using the at least one assignment rule.
7. The method of claim 4 , further comprising, after receiving the first plurality of data points and before selecting the similarity audience: performing the clustering algorithm using the first plurality of data points and the second plurality of data points to generate the plurality of clusters.
8. The method of claim 4 wherein the clustering algorithm is an iterative clustering algorithm that continues forming clusters until the number of unique visitors in respective clusters of the plurality of clusters is within a defined range of a total number of unique visitors.
9. The method of claim 8 wherein the number of unique visitors varies no more than 10%.
10. The method of claim 8 wherein the clustering algorithm is a spherical k-means clustering algorithm that associates at least some of the second plurality of visitors with more than one cluster of the plurality of clusters.
11. A method, comprising: receiving, for at least one website, a first plurality of data points related to each visitor of a first plurality of visitors to the website, the first plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; performing, using a computing device, a clustering algorithm that generates a plurality of clusters using the first plurality of data points as input, each cluster of the plurality of clusters defined by a number of unique visitors and at least some of the unique visitors sharing at least one interaction in common; generating, for each cluster of the plurality of clusters, at least one assignment rule that associates a behavior common to at least some of the unique visitors in the cluster with the cluster; providing the at least one assignment rule to an updating process that assigns each visitor of a second plurality of visitors to the plurality of clusters using the at least one assignment rule; and providing the plurality of clusters to a targeting process that generates a similarity audience formed of at least some of the clusters and provides, via electronic transmission, at least some members of the similarity audience with digital content.
12. The method of claim 11 , further comprising: receiving, for at least one website, a second plurality of data points related to each visitor of a third plurality of visitors to the website, the second plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; receiving a target audience comprising at least some of the third plurality of visitors having a known, desired interaction; receiving a plurality of selection rules defining tolerances for a similarity audience, the plurality of selection rules including a correlation score and an audience composition index; selecting the similarity audience from the plurality of clusters, the at least some of the plurality of clusters forming the similarity audience satisfying the plurality of selection rules; and generating, to a plurality of computing devices associated with the at least some members of the similarity audience, the digital content.
13. The method of claim 11 , further comprising: receiving, for at least one website, a second plurality of data points related to each visitor of a third plurality of visitors to the website, the second plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; assigning each visitor of the third plurality of visitors to at least one cluster of the plurality of clusters using the at least one assignment rule; re-assigning at least some visitors of the first plurality of visitors to at least one cluster of the plurality of clusters using the at least one assignment rule; determining whether to replace the plurality of clusters; and upon a determination to replace the plurality of clusters: performing the clustering algorithm a second time using the second plurality of data points and the first plurality of data points associated with the at least some visitors of the first plurality of visitors as input; and providing, to the targeting process, the plurality of clusters generated by performing the clustering algorithm the second time; otherwise: providing, to the targeting process, the plurality of clusters after the assigning and the re-assigning.
14. The method of claim 13 wherein determining whether to replace the plurality of clusters comprises at least one of: determining whether at least one of the plurality of clusters is larger than a desired cluster size; determining whether at least a defined period of time has passed between receiving the first plurality of data points and receiving the second plurality of data points; or determining whether at least one website exhibits a change in a taxonomy used to generate the plurality of clusters using the first plurality of data points.
15. An apparatus, comprising: a memory; and at least one processor configured to execute instructions stored in the memory to: receive, for at least one website, a first plurality of data points related to each visitor of a first plurality of visitors to the website, the first plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; receive a target audience comprising at least some of the first plurality of visitors having a known, desired interaction; receive, responsive to a user input, a plurality of selection rules defining tolerances for a similarity audience, the plurality of selection rules including a correlation score and an audience composition index; select the similarity audience from a plurality of clusters defined by a number of unique visitors, at least some of the unique visitors sharing at least one interaction in common, and the similarity audience comprising at least one cluster of the plurality of clusters satisfying the plurality of selection rules; and generate, for electronic transmission to a plurality of computing devices associated with members of the similarity audience, digital content.
16. The apparatus of claim 15 , the processor configured to: generate a cosine similarity index for each cluster of the plurality of clusters using the target audience as input; wherein the correlation score is a minimum value for the cosine similarity index of each cluster of the plurality of clusters forming the similarity audience.
17. The apparatus of claim 15 , the processor configured to: receive, before receiving the first plurality of data points and for at least one website, a second plurality of data points related to each visitor of a second plurality of visitors to the website, the second plurality of data points comprising at least an identification of the visitor and an interaction of the visitor with the website; and perform a clustering algorithm to generate the plurality of clusters using the second plurality of data points as input before receiving the first plurality of data points.
18. The apparatus of claim 17 , the processor configured to: perform, before selecting the similarity audience, the clustering algorithm to update the plurality of clusters using the first plurality of data points and at least some of the second plurality of data points as input.
19. The apparatus of claim 17 , the processor configured to: generate, for each cluster of the plurality of clusters, at least one assignment rule that associates a behavior common to at least some of the unique visitors in the cluster, with the cluster; assign each visitor of the first plurality of visitors to at least one cluster of the plurality of clusters, before selecting the similarity audience, using the assignment rules; and re-assign at least some visitors of the second plurality of visitors to at least one cluster of the plurality of clusters, before selecting the similarity audience, using the assignment rules.
20. The apparatus of claim 17 wherein the clustering algorithm is an iterative spherical k-means clustering algorithm that continues forming clusters until the number of unique visitors in each cluster of the plurality of clusters is varies no more than a defined percentage.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
October 17, 2016
November 20, 2018
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.